1. Field of the Invention
The present invention relates to a management system for monitoring a memory system placed in a network. More specifically, the present invention relates to a system which can freely access monitor information from a system connected to the network by eliminating a special function of the Operating System (OS) for accessing monitor information or a specific device driver to be used when the monitor information is accessed.
2. Description of the Related Arts
In case of managing disks which are shared between systems connected to a network, software dedicated to manage the shared disks is conventionally installed in the operating system of a HOST system, along with the device driver. For example, FIG. 18 shows the structure of a conventional data storage system. In FIG. 18, HOST system 2 has a dedicated device driver 2a for controlling a disk manager 1 and is connected to a data storage system 8 through a PCI BUS 11. The data storage system 8 includes shared disks composed of a Redundant Array of Inexpensive Disks (RAID). Storage disks 6 are the shared disks and disk manager 1 monitors and controls overall operation. Storage disks 6 are controlled by a disk controller 3, which is a RAID Control Board. The disk manager 1 collects the operation information for each storage disk from the disk controller 3 and storage disks 6. Then, the collected information is stored in a memory provided by the disk manager 1. FIG. 19 shows a detailed illustration of the memory. The device driver 2a dedicated to the control of the disk manager 1 is connected within the HOST system 2 in advance. When the operation information is obtained by the HOST system 2, the disk manager memory is accessed through the device driver 2a. The HOST system 2 instructs the device driver 2a to retrieve the operation information for the storage disks 6. Then, the device driver 2a retrieves the operation information for the storage disk 6 by accessing the memory of the disk manager 1. The HOST system 2 obtains the operation information for the storage disks 6 by receiving the operation information retrieved through the device driver 2a.
FIG. 19 shows an example of operation information stored by the disk manager. As shown in FIG. 19, the operation information is composed of ten blocks. Block Addresses 0 to 7 are blocks for Read and Block Addresses 8, 9 are blocks for Write. Block Addresses 0 to 9 are, small computer system interface (SCSI) Blocks. Each SCSI Block is accessible from the SCSI bus 12 and is stored in a static random stored access memory (SRAM) of the disk manager 1. Actually, data is stored in a Status Table in the SRAM of the disk manager 1. Operation information for each storage disk 6 and a fan 7 is transmitted through a Serial bus 14. The disk manager 1 then writes the transmitted operation information on the Status Table. In addition, operation information for a RAID System, RAID level information, and operation information for disks connected to the RAID system are transmitted to the disk manager 1 through a RAID system (RS) Port 13. The disk manager 1 writes the transmitted operation information in the Status Table. The disk manager 1 also extracts the requisite information requested by the inquiry from the device driver 2a provided by the HOST system 2, from Block Addresses 0 to 7, and transmits the information to the HOST system 2 through the device driver 2a. The status table may also be provided independent of the SCSI Blocks. In this case, the Status Table is pointed to by information stored in the SCSI Blocks.
A conventional disk manager 1 stores information from monitoring the operation of the storage and RAID system in the Status Table in the SRAM. However, the HOST system 2 needs a device driver 2a dedicated to the control of the disk manager to access the Status Table. Therefore, the following problem is caused.
When the operating system of the HOST system is changed, the device drivers are replaced. Then, modification of the device driver is required even when the device driver does not need to be replaced.
When the HOST system is connected by a network, a problem occurs because the monitor information is not able to be accessed directly from other systems connected to the network.
Also, when a fault occurs in a data storage system, data stored in the Status Table may be lost.
Furthermore, since the disk manager 1 has the Status Table, when the number of storage devices increase and the scale of the data storage system is enlarged, the size of the Status Table is enlarged and the load on the disk manager 1 increases.